Open xolod79 opened 2 months ago
How "movb" works on original LSI-11:
How "movb" works on M4:
How "movb" works on F-11:
Why should we change this behavior ? It seems it was intended to have the same as DEC did. Single write "movb also might have side effects on some I/O. Should this change be merged? Not sure, because of it impacts original DEC compatibility. Keeping the latter was the objective of BM2 developers.
VM2 byte write without RMW. In wb_plm.v simple change:
assign p[26] = cmp(rq[15:0], 16'b0001xxxxxxxxxxxx); to assign p[26] = cmp(rq[15:0], 16'bx001xxxxxxxxxxxx); // WTBT no RMW
Probably it is worth making a parameter during assembly for selection. Work is checked on DVK in RT11 and Soyuz Neon. Сorrection author Alexey Kisly.